Creontiades dilutus, the green mirid, is endemic to Australia and widely distributed across the continent. These bugs have been recorded on a broad range of host-plants including native species, weeds and several crops, particularly cotton, lucerne, and soy. The economic relevance of green mirids to the Australian cotton industry increased in recent years in response to the uptake of transgenic cotton, which controls Lepidopteran pests but is ineffective against Hemiptera. In this thesis I combined several molecular and ecological approaches to develop a better understanding of the species status of this insect, its use of multiple hosts and its long-distance movement.
Creontiades dilutus had reputedly been recorded in the USA during 2006. With collaborators at the USDA, I used sequence data (Cytochrome Oxidase I (COI) and 28S ribosomal gene) to establish that the insects concerned were highly unlikely to be C. dilutus. Subsequent taxonomic work confirmed that the USA species was indeed a separate species, Creontiades signatus. Using C. signatus as an out-group, further phylogenetic analyses showed that C. dilutus and C. pacificus are well differentiated according to the sequence of both genes. The COI sequences also indicated low levels of genetic diversity in C. dilutus (Pi = 0.0006), especially in comparison to C. pacificus (Pi = 0.0026). The low COI diversity indicated that more variable markers would be required for further analyses of gene flow in this species, and consequently 12 microsatellites were developed by enrichment.
To understand the use of multiple hosts by C. dilutus, all the available host plant data were analysed. Most of the putative host plants recorded prior to this thesis were crop species or introduced weeds. As C. dilutus has not been recorded outside of Australia it was evident that a more thorough investigation of potential native hosts was necessary. Over three seasons of field surveys in central and eastern Australia I added an additional 25 species to the list of potential hosts, 22 of which are native to Australia. The presence of nymphs indicates that C. dilutus is indeed able to feed and reproduce on 46 host plant species, most in the family Fabaceae. Quantitative sampling, however, revealed a strong association between C. dilutus and two plant species in the genus Cullen. These two species are thus likely the primary host plants for green mirids. To test whether green mirid individuals show a strong preference for Cullen under field conditions I amplified Chloroplast DNA from DNA extracted from whole insects. These diet analyses demonstrated that C. dilutus individuals do feed on alternative host plants to the one from which they were collected, even when that was a Cullen species, and multiple host use by individuals was not infrequent. Green mirids are found, sometimes in large numbers, in arid parts of central Australia, and Miles (1995) suggested that this might be the main source of the mirids invading cotton and other crops in sub coastal eastern Australia.
Green mirid abundance is seasonally inverse between the central arid regions and eastern cropping areas, and they likely experience different selective pressures in each region. The population genetic consequences of these dynamics were assessed by sequencing a mitochondrial COI fragment from individuals collected over 24 years, and screening microsatellite variation for 32 populations across two seasons. A single COI haplotype predominated in samples from 2006/2007, but in the older collections (1983 and 1993) a different haplotype was most prevalent. This is consistent with successive population contractions and expansions, likely in response to alternate periods of drought and flood in the arid interior of Australia. The microsatellite data showed genetic differentiation between population samples, evidence for movement between sites, and also genetic signatures of bottleneck events. The Simpson Desert, in central Australia was identified as a source of recent immigrants to populations in Biloela (m = 0.15, BAYESASS), eastern Australia, supporting the view that long-distance migration is, indeed, a regular part of the ecology of this species. Together, these data highlight that since the advent of agriculture in Australia, green mirid dynamics are still shaped by its adaptations to arid, spatiotemporally variable environments.
Previous ecological studies presented evidence that C. dilutus may be a complex of cryptic species, and that two such cryptic species may be associated with cotton and lucerne crop hosts. Further, C. dilutus has a reported preference for lucerne over cotton, leading to the proposal of lucerne as a trap crop in cotton production systems. To test this proposition I sampled C. dilutus individuals from adjacent cotton and lucerne crops at three geographically separate sites within a single season. Individual-based clustering analyses using microsatellite data showed that gene flow was high across these crop hosts. Further, gut content analysis indicated that a relatively high proportion of individuals collected from one crop host had fed on the alternative host and several individuals had fed on both. These data support the presence of one species associated with cotton and lucerne, but also show that green mirids will readily move between these two hosts despite their relative preference for lucerne.
The findings outlined above are discussed in relation to ecological perceptions of generalist habits, the application of genetic techniques to the solution of ecological problems involving multiple host use, and the management and research implications arising from the data presented in this thesis.